Update on the AOP's engagement survey

On Wednesday the AOP released the results of a survey on "the relationship between consumer engagement with website content and how it impacts on their reaction to advertising on different website types: original content sites, portals and social networks."It found that approximately twice as many respondents to the survey trusted the advertising on content sites as on social media

It raised some questions for me which I posted here. If trust and engagement are such key indicators of the value of online advertising, how can content sites maintain CPMs that are ten times greater than those achieved by social media when their own research is able to show only twice as much trust for their advertising?

AOP came back to me yesterday afternoon (the delay in posting their response only now is therefore mine). Most of my detailed questions were about the surveys methodology, and I am satisfied by their answers that the survey is robust and fair. On the implications for revenues, they comment

"We can’t talk about individual pricing differentials, these are affected by many factors including category, format, positioning etc. Value itself is not entirely related to trust there are many variables to consider, it is too simplistic an approach to merely compare the levels of trust with pricing mechanisms."

Which of course is true, as far as it goes - the value of advertising is determined by many factors, not only trust - but do any of the factors cited, taken individually or collectively, account for the apparent ability of content sites to charge ten times as much for advertising as social media?

Format and positioning are absolutely trivial to replicate. From time to time new and more effective ad formats emerge and when they do all of the ad-funded sites take a look at them and, if they seem to work, move over to using them. If Facebook can move from its current half dollar CPM even appreciably closer to the five plus dollars CPM enjoyed by content publishers merely by switching from (say) MPUs below the fold to banners above it that's massive news for the display ad market and information Facebook's expanded UK ad sales team should really be able to use.

Category, by contrast, is specific to content sites and it would be non-trivial for Facebook to build eg finance, current affairs or travel pages. Since high-yielding category advertisers usually prefer to appear in relevant sections this is a good reason for content sites to charge more for their space, at least until Facebook works out a way of behaviourally targeting ads to people who are talking about travel plans or their finances.

So I can believe that category provides some uplift in value to the advertiser. I can believe that trust does also. But still, taken together, a sustainable tenfold uplift? That I am still having trouble believing, and it seems obvious, at least on the basis of our current knowledge, that content site CPMs and Facebook's CPMs are going to move a lot closer together in the coming months and years.